Top down market sizing starts with a broad number, usually a population or total category revenue figure, and narrows it through a series of filters until you land on TAM, then SAM, then SOM. It works best for consumer-facing or demand-driven markets, quick estimates, and case interviews where structure matters more than precision. The rest of this piece walks through the exact steps, a full worked example, and the traps that trip up most people who try it.
TL;DR:
- Top-down sizing provides quick, directional estimates but relies heavily on the accuracy of large, stable base figures and transparent filters.
- Applying demographic and behavioral filters step-by-step, with clear assumptions, improves the defensibility of your market size and reveals sensitivity to key variables.
- Combining top-down with bottom-up approaches offers more reliable insights, especially when early validation like landing pages or interviews confirms assumptions.
- Using structured tools like Klaritea links market sizing to business planning, automates updates, and challenges assumptions with AI, reducing errors and improving documentation.
- In early stages, top-down is most useful for ruling out markets; precise economic metrics require primary research, such as customer interviews or limited testing.
Table of Contents
- What top-down market sizing is and how it differs from bottom-up
- A five-step method you can use in interviews or on the job
- Worked example: sizing an at-home fitness app market
- Common mistakes and how to present your assumptions
- How Klaritea's founder came to build a market sizing tool
- When top-down helped and when it didn't
- Running your market sizing inside Klaritea instead of a spreadsheet
- Curated sources and further reading
- FAQ
What top-down market sizing is and how it differs from bottom-up
Top down market sizing works backward from the whole. You pick a large, well-documented figure such as a national population, a total industry revenue number, or a category spend estimate, then apply a sequence of narrowing filters (who needs this, who will pay, how much will they pay) until you reach a usable estimate. The method trades precision for speed: you can build a defensible number in minutes because the starting figure is usually public and stable.
Bottom up market sizing runs the opposite direction. You start from a single unit, such as one customer, one store, or one sales rep's output, and multiply that unit economics figure by however many units you expect to sell or serve. Bottom up tends to be more accurate when you have real data on pricing, conversion, or unit costs, but it takes longer and depends on assumptions that are often harder to source than a national population figure.
Choosing between them comes down to what you have and what you need:
- Use top down when you need a fast, directional estimate and the market is demand-driven, meaning total demand is roughly bounded by population or existing category spend.
- Use bottom up when you have specific data on your own funnel, pricing, or comparable unit economics and need a tighter number.
- Combine both when the decision matters enough to justify triangulating two independent paths to the same figure.
In practice, top down maps cleanly onto the standard TAM, SAM, and SOM framework. TAM (total addressable market) is your broad base figure after applying the widest relevant filter. SAM (serviceable addressable market) narrows that to the segment your product or business model can realistically serve, given geography, channel, or product fit constraints. SOM (serviceable obtainable market) narrows again to the share you can realistically capture given competition, budget, and time. Klaritea's TAM/SAM/SOM guide walks through this progression with templates for converting population or user estimates into revenue.
The quality of a top-down estimate depends heavily on which base figure you choose and how sensitive your final number is to small changes in your filter percentages. A market sizing analysis built on a shaky base number, or one that stacks several rough guesses on top of each other, can drift far from reality even when every individual step looks reasonable. Picking a well-sourced base figure and stress-testing your filters against a plausible range, rather than a single point estimate, is what separates a useful estimate from a guess dressed up in a spreadsheet.
A five-step method you can use in interviews or on the job
This is the version of top down market sizing that holds up whether you're answering a case interview prompt in ninety seconds or building a real model for a pitch deck. It has five steps, each with a clear output, so you always know whether you're ready to move to the next one.
- Define the market boundary. State exactly who and what you're sizing: geography, customer type, and product category. A vague boundary is the single biggest cause of a wrong answer.
- Choose a base figure. Pick a large, citable number such as total population, total households, or total category revenue that plausibly bounds your market from above.
- Apply demographic or behavioral filters. Narrow the base step by step using relevant percentages: age range, income band, existing behavior, or willingness to adopt a new category.
- Convert to revenue or units. Multiply your filtered population by a price, average revenue per user, or units-per-year figure to turn a headcount into a dollar or unit estimate.
- Sanity-check and present your assumptions. Compare your result against any public benchmark you know, and state your filters out loud (or in writing) so a reviewer can follow your logic.
Base data for step 2 ideally comes from a national statistics agency, an industry association report, or a market research firm's published category estimate. In an interview setting where you can't look anything up, reasonable substitutes are round population figures you already know (a country's population, average household size) and category assumptions the interviewer confirms or lets you state as a working assumption. The goal is a defensible number, not a memorized one.
For step 3, state each percentage as you apply it rather than burying several filters into one multiplication. If you're estimating adoption, frequency, and conversion in the same problem, apply them one at a time: "of that group, I'll assume 30% have some interest in the category, and of those, maybe 10% would actually pay for a premium version." Round aggressively. Precision beyond one or two significant figures is wasted effort in a market sizing exercise, since your inputs are estimates in the first place.
Time management separates people who look confident from people who look like they're guessing. In the first 60 to 90 seconds, restate the problem, define your boundary, and name your base figure and your planned filters out loud, before doing any math. This signals structure early and gives the interviewer a chance to redirect you if you've misunderstood the prompt. Spend the remaining time working through the filters in order, narrating each step, and finish with a one-line sanity check comparing your number to something you know (a comparable company's reported revenue, a rough sense of category size).
Pro Tip: Write your five steps as labeled lines on paper or a whiteboard before you touch a single number. It keeps you from losing your place and makes your logic easy for someone else to follow.
Worked example: sizing an at-home fitness app market
Here's a full pass through the method using a common interview prompt: estimate annual revenue for a subscription-based at-home fitness app in a large-market country of roughly 60 million adults. Before any math, state the assumptions plainly, since a market sizing analysis lives or dies on whether its assumptions are visible and reasonable.
Assumptions for this scenario:
- Adult population: 60 million (a stated, round figure for the exercise)
- Smartphone or connected-device ownership among adults: 85%
- Share of device owners interested in a paid fitness app: 8%
- Conversion from interest to paying subscriber: 40%
- Average revenue per paying user (ARPU): $120 per year
Applying a 40% conversion rate from interest to paying subscriber gives roughly 1.63 million paying users. Multiplying by $120 ARPU gives an estimated annual revenue of about $196 million.
Rounded for an interview setting, you'd present this as "roughly $200 million a year," and that rounding is the right call. Precision to the dollar implies false confidence in inputs that are themselves educated guesses.
A second scenario with a more conservative adoption assumption shows how sensitive the estimate is to your filters. That's a meaningful swing from a single filter change, which is exactly why stating assumptions clearly, and being ready to show a range rather than one number, matters more than nailing a precise figure.
A reasonable top-down estimate for a national at-home fitness app market is in the low hundreds of millions of dollars annually, and the exercise's value lies in showing how each filter moves that number, not in hitting a single "correct" answer. Sanity-checking against a known comparable, such as a public fitness or wellness app's reported user base or revenue range, is the last step: if your estimate is off by an order of magnitude from anything comparable you know, revisit your filters before presenting the number.
Common mistakes and how to present your assumptions
Most bad top-down estimates fail for a small number of repeatable reasons, and most of them are fixable once you know to watch for them.
- Wrong base population. Sizing a product for "adults" when the real addressable group is "urban adults with disposable income" inflates every downstream number.
- Mixing incompatible rates. Applying an annual adoption rate to a monthly revenue figure, or vice versa, without converting units first.
- Double-counting. Applying two filters that overlap, such as "smartphone owners" and "app store users," as if they were independent populations.
- Mis-applied unit conversions. Forgetting to convert household figures to individual figures, or annual figures to monthly ones, partway through a calculation.
When you present assumptions, say them as short, specific statements rather than long qualified sentences: "I'm assuming 60 million adults, 85% device penetration, 8% category interest." A reviewer can challenge a crisp assumption far more easily than a paragraph of hedging, and that's actually good, since a challenged assumption you can defend is worth more than one nobody questioned.
On accuracy: in most interview and early-planning contexts, landing within half to double of a reasonable benchmark is considered acceptable, because the goal is defensible structure, not decimal precision. Interview-prep guidance consistently emphasizes that the reasoning process, not the final digit, is what's being evaluated.
Pro Tip: Keep a short mental list of round numbers you can reach for under pressure: a country's approximate population, rough global internet penetration, typical smartphone ownership rates in developed markets. Having these ready saves time you'd otherwise spend guessing on the spot.
How Klaritea's founder came to build a market sizing tool
Karl built Klaritea after watching founders repeatedly skip or botch the market sizing step before writing a line of code, treating TAM/SAM/SOM as a slide to fill in for investors rather than a real constraint on the business. That gap between what founders assumed about their market and what a rigorous top-down pass would have shown became the starting point for the product.
There is an AI-powered "phase 0" planning tool that turns a one-line idea into a structured, connected model of an entire business, including market sizing (ICP, TAM/SAM/SOM), competitor analysis, features, requirements, a build specification, and a pitch, all linked so a change in one area updates the rest. Inside the platform, an AI advisory board of three specialized advisors researches, challenges, and fact-checks the assumptions a founder enters, which is the same discipline this article recommends applying manually.
A spreadsheet is fine when you're sizing one market once. A structured tool earns its cost when you need to revisit assumptions repeatedly, defend them to investors, or keep market sizing synchronized with a changing product plan.
When top-down helped and when it didn't
Top-down estimates are most useful early, when you need a directional number fast to decide whether an idea is worth pursuing further. A rough TAM built from population and category filters can rule out a market that's simply too small before you spend weeks on anything else.
Where top-down falls short is precision on your own unit economics. It can't tell you your actual conversion rate, your actual churn, or your actual price sensitivity, and treating a top-down SOM as a real revenue forecast is a common overreach. The fix is cheap: run a small landing page test, a handful of customer interviews, or a limited pre-sale before committing real budget. Move from estimation to primary research as soon as the decision in front of you is expensive enough that being wrong costs more than the research would.
— Karl
Running your market sizing inside Klaritea instead of a spreadsheet
Klaritea keeps the whole model connected, so a change to your population assumption or ARPU flows through to every downstream number and every linked section of your business plan automatically.
- A connected model links your TAM/SAM/SOM directly to your ICP, features, and requirements, so market sizing isn't an isolated slide.
- The AI advisory board challenges your filter assumptions the way a sharp reviewer would, before you present them to anyone else.
- Exportable, printable reports and a build spec turn your market sizing work into something you can actually hand to an investor or a developer.
- Templates for TAM/SAM/SOM keep your assumptions traceable instead of buried in cell references.
If you're a founder or analyst who wants a documented, reproducible model instead of a spreadsheet held together by memory, Klaritea's pricing page lists the Free, Klaritea, and Pro plans, along with credit packs starting at 750 credits for $15, so you can see which tier fits before you commit.
Curated sources and further reading
For templates and a deeper walkthrough of converting population estimates into revenue, Klaritea's TAM/SAM/SOM guide covers the full progression with worked examples. For finding reliable base data such as census figures and industry reports, Klaritea's roundup of market research tools points to practical starting places. For a look at applying top-down thinking specifically to SaaS and technology markets, this SEO strategy guide includes examples of narrowing broad category revenue down to a serviceable segment.
FAQ
What is top-down and bottom-up market sizing?
Top-down market sizing starts from a broad figure, such as total population or category revenue, and narrows it through filters to reach TAM, SAM, and SOM. Bottom-up market sizing starts from a single unit, like one customer's spend, and multiplies it by an expected number of units to build up to a total.
What is TAM and SAM and SOM?
TAM is the total addressable market, the full demand for a product category if you served everyone who could plausibly want it. SAM is the serviceable addressable market, the slice you could realistically serve given your business model and geography, and SOM is the serviceable obtainable market, the share you could realistically capture given competition and resources.
Can you give an example of a market sizing exercise?
A common example is estimating annual revenue for an at-home fitness app by starting with a country's adult population, applying filters for device ownership, category interest, and paid conversion, then multiplying the resulting paying users by an average revenue per user. Working through such filters typically produces an estimate in the low hundreds of millions of dollars for a market of that scale, rounded for presentation rather than stated to the exact dollar.
What is a good TAM?
A good TAM is one built on a clearly stated, defensible base figure and transparent filters, not necessarily a large number. What matters more than the size itself is whether your SAM and SOM show a credible, obtainable path to revenue once you've narrowed from that TAM.